Trail Making Test: Normative data for Turkish elderly population by age, sex and education

Abstract

Objective

Trail Making Test (TMT) is a neuropsychological test, which has parts A and B that can precisely measure executive functions, like complex visual-motor conceptual screening, planning, organization, abstract thinking and response inhibition. The main purpose of this study is to standardize TMT for Turkish adults and/or elderly population. This study primarily consists of two main parts; norm determination study and reliability/validity studies, respectively.

Methods

The standardization study was carried on 484 participants (238 female and 246 male). Participants at the age of 50 years and older were selected from a pool of people employed in or retired from governmental and/or private institutions. The research design of this study involves the following variables mainly; age (7 subgroups), sex (2 subgroups) and education (3 subgroups).

Results

Age, sex and education variables have significant influence on eight different kinds of TMT scores. Statistical analysis by ANOVA revealed a major effect of age (p < 0.001) and education (p < 0.001) on time spent in Part A or B, or time difference between Parts B and A, or sum of Parts A and B. Similarly, influence of sex (p < 0.05) on time spent on Part A or B, or sum of Parts A and B was shown to be significant. Kruskal–Wallis Test was performed and chi-square (χ²) values revealed that, correction scores for Part A and B were found to be influenced by age groups (p < 0.001). Test–retest reliability and inter-rater reliability coefficients for time scores of Parts A and B were estimated as 0.78, 0.99 and 0.73, 0.93, respectively.

Conclusion

This study provides normative data for a psychometric tool that reliably measures the executive functions in Turkish elderly population at the age of 50 and over.

Introduction

Trail Making Test (TMT) was first developed in 1944 as a visual–motor and visual–conceptual trail making test, as a part of ‘Army Individual Test Battery’, which was prepared by psychologists in the United States Army and released for public use in the following years [1]. The first version of TMT was composed of two separate parts, as A and B. In the first part (Part A), the participant is required to combine the numerated circles with direct lines to provide correct alignment of numbers in an ascending order (as 1-2-3-4-…). In the second part (Part B), the participant is required to combine the circles, which are consisted of both letters and numbers with direct lines to provide correct alignment of both numbers and letters in an ascending order, as one number and one letter (1-A-2-B-3-C-…).

Test procedure and scoring was subjected to a number of modifications in time during its development. In its original version, the test was completed by the tester after three corrected mistakes. Every trial was assessed on a 10 point scale and the length of time to complete the test comprised the primary data for the assessment. Armigate et al. modified this assessment method without considering the number of mistakes they had made and permitted the participants to continue the test even though three mistakes or more were recorded [2]. Corrected mistakes were recorded as zero points. Similarly, Reitan et al. ignored the number of mistakes and measured the time period for test completion [3]. In the following years, Spreen and Staruss prepared detailed instructive measures for test practice and restricted the time period (4 to 5 min) necessary to complete the test [4]. Currently, simplified TMT scoring method is still being used in the form proposed by Reitan et al., despite some critics [3]. Some authors criticized these instructive measures, which seemed to threaten the test reliability, depending on the fact that it could result in some mistakes due to reaction time of the tester, if the duration of test is regarded. Moreover, additional mistakes could be made if a participant tries to be faster and more time is spent for correcting the mistakes. This method indirectly punishes the mistakes, however does not control the duration of time spent for mistakes and/or corrections. As a summary, there is a potential bias on the time scoring method due to individual differences [5], [6].

The difference obtained by subtraction of Part A score from Part B score (Score B − A) is used to eliminate speed component, which is a variable in evaluation of TMT. This difference is shown to be highly correlated with both mental ability tests and various cognitive disturbances [7]. The TMT test was further developed into different forms, like Oral Trail Making Test [8] and Comprehensive Trail Making Test [9], which were adapted to various cultures and/or languages [10], [11].

TMT is one of the most widely used screening tests in clinical neuropsychology practice [4], [6], [9]. From neuropsychological point of view, frontal lobes are brain regions responsible from the higher levels of cognitive functions [6]. Working memory, complex attention, planning, cognitive set shifting, response inhibition, problem solving and other executive functions may well be taken into account when the role of frontal lobe is considered. TMT, particularly Part B, is a widely used neuropsychological scale for executive functions like complex attention, planning, cognitive set shifting and response inhibition [6], [12]. Participants with frontal lobe damage were shown to have low Part B scores, while no difference was detected among the Part A scores of patients with or without any frontal lobe damage [12], [13]. Nevertheless, there are other clinical studies featuring patients with frontal lobe damage with low performance scores both in Parts A and B in comparison to normal participants [14], [15].

In a particular study concerning patients with mild, moderate and severe dementia, patients were compared in terms of Part A, Part B and total time scores of Parts A and B (Score A + B). When the dementia increases in severity, mean time spent to complete each part increased and the difference between groups was found to be statistically significant [16]. Besides, association between TMT time scores and Wechsler Adult Intelligence Scores (WAIS) was investigated, and TMT time scores of Parts A and B were found to be inversely correlated with WAIS scores (verbal and performance subtest scores, together with total IQ scores) [16], [17].